광역조사 라만 분광법을 이용한 신뢰성 있는 고체 과립 시료 분석

Abstract

The utility of wide area illumination (WAI) Raman scheme for the analysis of granular samples has been demonstrated in this thesis. Since the sizes of granules are fairly large (3 - 7 mm), correct sample representation should be the major requirement for reliable Raman spectroscopic analysis. First, I WAI Raman scheme was used to differentiate rices of two different geographical origins. The rice samples used were whole grains that were not further ground. For comparison, both the WAI and conventional Raman schemes were used to collect spectra of all grains. The WAI scheme exhibited improved differentiation, primarily due to the fact that the WAI scheme could efficiently produce Raman spectra with a more reliable sample representation, as well as better reproducibility. The spectral feature obtained using the conventional scheme appeared to be more variable; however, this variation resulted more from unsuccessful sample representation rather than from an actual change in the sample composition. Second, the sample representation of WAI scheme has been evaluated by comparing both spectra of polyethylene (PE) pellets collected with and without sample rotation. When the pellet samples were not rotated during spectral collection, the resulting Raman intensities were quite changed due to the variation of sample packing at each measurement with respective to the size of laser illumination area; however, overall spectral features were nearly same as those obtained by rotating pellet samples. The intensity variation induced by irregular sample packing was effectively minimized with the use of normalization. Partial least squares (PLS) prediction accuracies for the determination of density were almost identical whether samples were rotated or not during spectral collection.